INTRODUCTION
⌅Batoidea infraclass (hereafter “batoids”), which include skates, rays, stingrays and their allies (Carrier et al. 2004Carrier J.C., Musick J.A., Heithaus M.R. 2004. Biology of sharks and their relatives. CRC Press. https://doi.org/10.1201/9780203491317 ), are considered vulnerable species (Dulvy et al. 2014Dulvy N.K., Fowler S.L., Musick J.A., et al. 2014. Extinction risk and conservation of the world’s sharks and rays. Elife 3: e00590. https://doi.org/10.7554/eLife.00590 ). In general, the high vulnerability of batoids in comparison with most teleost fishes is related to vital traits such as their large body size, slow growth, late sexual maturation and low fecundity (Stevens et al. 2000Stevens J.D., Bonfil R., Dulvy N.K., Walker P.A. 2000. The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J. Mar. Sci. 57: 476-494. https://doi.org/10.1006/jmsc.2000.0724 , Dulvy and Reynolds 2002Dulvy N.K., Reynolds J.D. 2002. Predicting extinction vulnerability in skates. Conserv. Biol. 16: 440-450. https://doi.org/10.1046/j.1523-1739.2002.00416.x , Frisk et al. 2005Frisk M., Miller T.J., Dulvy, N.K. 2005. Life Histories and Vulnerability to Exploitation of Elasmobranchs: Inferences from Elasticity, Perturbation and Phylogenetic Analyses. J. Northw. Atl. Fish. Sci. 35: 27-45. https://doi.org/10.2960/J.v35.m514 ). Although batoids are not usually target but by-catch species, these vital traits make them particularly vulnerable to the impacts of fishing exploitation, resulting in a global decline and a large proportion of species being assessed as overexploited (Stevens et al. 2000Stevens J.D., Bonfil R., Dulvy N.K., Walker P.A. 2000. The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J. Mar. Sci. 57: 476-494. https://doi.org/10.1006/jmsc.2000.0724 , Worm et al. 2013Worm B., Davis B., Kettemer L., et al. 2013. Global catches, exploitation rates, and rebuilding options for sharks. Mar. Policy. 40: 194-204. https://doi.org/10.1016/j.marpol.2012.12.034 , Dulvy et al. 2014Dulvy N.K., Fowler S.L., Musick J.A., et al. 2014. Extinction risk and conservation of the world’s sharks and rays. Elife 3: e00590. https://doi.org/10.7554/eLife.00590 ).
The Mediterranean Sea has been reported as one of the regions where elasmobranchs are most threatened in the world (Dulvy et al. 2014Dulvy N.K., Fowler S.L., Musick J.A., et al. 2014. Extinction risk and conservation of the world’s sharks and rays. Elife 3: e00590. https://doi.org/10.7554/eLife.00590 ). In this sea, up to 15 species of skates (Rajidae) are present (Meléndez et al. 2017Meléndez M.J., Báez J.C., Serna-Quintero J.M., et al. 2017. Historical and ecological drivers of the spatial pattern of Chondrichthyes species richness in the Mediterranean Sea. PLoS ONE 12: e0175699. https://doi.org/10.1371/journal.pone.0175699 ), three of them considered endemic: Leucoraja melitensis, Raja radula and Raja polystigma (Dulvy et al. 2016Dulvy N.K., Allen D.J., Ralph G.M., Walls R.H.L. 2016. The conservation status of sharks, rays and chimaeras in the Mediterranean Sea (Brochure). IUCN, Malaga, Spain.). Raja asterias, formerly considered a Mediterranean endemism, was recently reported from the gulf of Cádiz, northeastern Atlantic (Ordines et al. 2017Ordines F., Baro J., Ramírez-Amaro S., et al. 2017. First substantiated record of Raja asterias Delaroche, 1809 (Elasmobranchii: Rajiformes: Rajidae) in the Gulf of Cádiz, north-eastern Atlantic. Acta Ichthyol. Piscat. 47: 101-106. https://doi.org/10.3750/AIEP/02161 ).
The speckled ray, Raja polystigma (Regan, 1923), is a small skate, with a maximum total length of about 60 cm (Stehmann and Bürkel 1984Stehmann M., Bürkel D.L. 1984. Rajidae. In: Whitehead P. J. P., Bauchot M. L., Hureau J. C., Nielsen J., Tortonese E. (eds), Fishes of the North-eastern Atlantic and Mediterranean. UNESCO, Paris, pp. 163-196., Frodella et al. 2016Frodella N., Cannas R., Velonà A., et al. 2016. Population connectivity and phylogeography of the Mediterranean endemic skate Raja polystigma and evidence of its hybridization with the parapatric sibling R. montagui. Mar. Ecol. Prog. Ser. 554: 99-113. https://doi.org/10.3354/meps11799 ). It is found in soft-bottom habitats ranging from 100 to 400 m depth and has been reported from the #northwest Iberian Peninsula, the Gulf of Lion, the Ligurian and Tyrrhenian seas, the Tunisian and Algerian coasts, the Adriatic Sea and the Greek coast (Capapé and Quignard 1978Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244., Capapé et al. 2006Capapé C., Guélorget O., Vergne Y., et al. 2006. Skates and rays (Chondrichthyes) from waters off the Languedocian coast (southern France, northern Mediterranean): a historical survey and present status. Ann. Ser. Hist. Nat. 16: 165-178.). It can be an abundant by-catch of the bottom trawl fisheries in some areas of the northwestern Mediterranean, such as the Balearic Islands (Ramírez-Amaro et al. 2015Ramírez-Amaro S., Ordines F., Terrasa B., et al. 2015. Demersal chondrichthyans in the western Mediterranean: assemblages and biological parameters of their main species. Mar. Freshw. Res. 67: 636-652. https://doi.org/10.1071/MF15093 , Follesa et al. 2019Follesa M.C., Marongiu M.F., Zupa W., et al. 2019. Spatial variability of Chondrichthyes in the northern Mediterranean. Sci. Mar. 83: 81-100. https://doi.org/10.3989/scimar.04998.23A ) and Sardinia (Porcu et al. 2020Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 ), although it can be less abundant or rare in other areas, such as the Strait of Sicily (Geraci et al. 2021Geraci M.L., Ragonese S., Scannella D., et al. 2021. Batoid abundances, spatial distribution, and life history traits in the strait of Sicily (Central Mediterranean Sea): Bridging a knowledge gap through three decades of survey. Animals 11: 2189. https://doi.org/10.3390/ani11082189 ). According to the International Union of Conservation of Nature (IUCN), this species is listed as least concern, although its population has an unknown trend (Ungaro et al. 2015Ungaro N., Dulvy N.K., Tinti F., et al. 2015. Raja polystigma. The IUCN Red List of Threatened Species.). The high intraspecific coloration variability, which makes it difficult to identify, and the confusion with Raja montagui (Cannas et al. 2008Cannas R., Pasolini P., Mancusi C., et al. 2008. Distribution, molecular systematics and phylogeography of Raja polystigma and Raja montagui in the Mediterraneanean. Biol. Mar. Mediterr. 15: 188-191., Ramírez-Amaro et al. 2018Ramírez-Amaro S., Ordines F., Picornell A., et al. 2018. The evolutionary history of Mediterranean Batoidea (Chondrichthyes: Neoselachii). Zool. Scr. 47: 686-698. https://doi.org/10.1111/zsc.12315 ) may have prevented the analysis of this trend (Frodella et al. 2016Frodella N., Cannas R., Velonà A., et al. 2016. Population connectivity and phylogeography of the Mediterranean endemic skate Raja polystigma and evidence of its hybridization with the parapatric sibling R. montagui. Mar. Ecol. Prog. Ser. 554: 99-113. https://doi.org/10.3354/meps11799 ).
The reproduction of R. polystigma seems to show a large geographic variability in both seasonality and attainment of sexual maturity, even at medium spatial scales. Capapé and Quignard (1978)Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244. reported winter as the season with the highest spawning activity on the Tunisian coast, with spawning females (with presence of egg-cases in the oviducts) from 35 cm in total length (TL), with more than 80% of spawning females above 39 cm TL. Porcu et al. (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 reported spring and summer as the spawning season and the smaller spawning females above 50 cm TL in Sardinia.
So far there is no complete biological and ecological description of the species in any of the areas where it has been studied. These data are needed for many stock assessment models that allow the state of exploitation of the resources to be evaluated and are the basis of a suitable management of fisheries. Moreover, assessments are scarce for Mediterranean batoids in general, including R. polystigma, mainly because of their low economic value and the monitoring programmes aimed at collecting the data needed for stock assessment, which prioritize the populations of target species (Ungaro et al. 2015Ungaro N., Dulvy N.K., Tinti F., et al. 2015. Raja polystigma. The IUCN Red List of Threatened Species.).
The aim of the present work is to study the main biological characteristics (population structure, distribution, condition, reproduction and diet) of R. polystigma in the Balearic Islands and to assess, for the first time, the trend of its population using the data collected during the time series of the MEDITS survey in this area.
MATERIALS AND METHODS
⌅Data source
⌅Information was obtained from two different sources: the monitoring of the commercial fleet in the European Data Collection Framework and the International Bottom Trawl Survey in the Mediterranean (MEDITS).
Commercial catches were collected by an observer on board the commercial bottom trawl fleet from Mallorca between May 2017 and February 2019. Samples were collected from four different harbours of Mallorca on the northwest (Sóller), northeast (Alcúdia and Cala Rajada) and southeast (Cala Figuera) coasts of Mallorca. This trawling fleet uses a square-mesh codend with a mesh size of 40 mm. During each commercial haul, duration, position, depth and vessel speed were recorded. These hauls were performed between 50 and 170 m depth. All the individuals of R. polystigma were transported to the laboratory of the Centre Oceanogràfic de Balears (COB), stored at -20°C and sampled biologically on a monthly basis. This biological sampling obtained in these individuals was as follows: individual TL and disc width (DW), both to the lowest mm, and total (TW), eviscerated (EW), liver (LW), digestive tract (SW) and gonad (GW) weights (±0.1 g); GW includes the ovaries in females and the testes in males. In the case of females with vitellogenic oocytes, these were counted and measured to the lowest 0.1 mm. Egg capsules were also counted when present. Stomachs were used for diet analysis.
Sex and maturity stage were determined macroscopically following a maturity scale (Table 1) based on the description of maturity stages in Brown-Peterson et al. (2011)Brown-Peterson N.J., Wyanski D.M., Saborido-Rey F., et al. 2011. A Standardized Terminology for Describing Reproductive Development in Fishes. Mar. Coast. Fish. 3: 52-70. https://doi.org/10.1080/19425120.2011.555724 but adapted to the studied species. The commercial catch data were used for the following sections: population structure, reproduction, condition indices (evolution throughout the year is needed, as well as some of the biological information, e.g. EW, LW and SW, which is not available in the surveys’ data), allometric growth (DW is needed but is not available in the surveys’ data) and diet.
| Maturity stage | Sex | Description |
|---|---|---|
| Immature | F | Absence of visible follicles. Oviductal glands not detected. |
| M | Non-solidified claspers, which do not reach the pelvic fin. Testes small or not distinct. Sperm not present in the sperm ducts. | |
| Developing | F | Large vitellogenic oocytes not detected. Visible oviductal glands. |
| M | Claspers reach or exceed the pelvic fin, but are not completely solidified. Testes are detectable but not forming sperm. | |
| Spawning | F | Large vitellogenic oocytes detected. |
| M | Claspers reach beyond the pelvic fin and are completely solidified. Presence of sperm in the sperm ducts. | |
| Capsule present | F | Large vitellogenic oocytes detected along with the presence of capsules. |
| Regenerating | F | Large vitellogenic oocytes not detected. Flaccid and dilated ducts and uterus. |
MEDITS surveys take place annually in late spring or early summer. In the present study, data and samples were collected between the years 2003 and 2018 in Geographical Subarea (GSA) 5 during these surveys. These data were obtained following the international MEDITS protocol. (For specific information on the sampling scheme see Bertrand et at. 2002Bertrand J.A., Gil de Sola L., Papaconstantinou C., et al. 2002. The general specifications of the MEDITS surveys. Sci. Mar. 74: 9-17. https://doi.org/10.3989/scimar.2002.66s29 , Spedicato et al. 2019Spedicato M.T., Massutí E., Mérigot B., et al. 2019. The MEDITS trawl survey specifications in an ecosystem approach to fishery management. Sci. Mar. 83S1: 9-20. https://doi.org/10.3989/scimar.04915.11X ). During the surveys, an experimental bottom trawl (GOC 73) with a codend mesh size of 18 mm (stretched mesh) is used to sample stations between 50 and 800 m depth. At each station, time, speed, position, depth and horizontal and vertical openings of the net during the haul (averaging 2.5 and 18 m, respectively) are recorded using a GPS and a SCANMAR system, respectively. Once the samples are on board, catches are sorted out by species, counted, sexed and weighed, and the individuals are measured (by TL in the case of R. polystgima). In the case of certain species, as R. polystigma, the stomach content analysis is also carried out. Density and biomass were standardized to 1 km2 using the swept surface during each haul, calculated using the distance covered through the bottom and the horizontal opening of the net. These data were used for geographic and bathymetric distribution, population trend and diet. The only section using data from both sources was diet analysis.
Population structure
⌅The size distribution for TL was obtained by season and for the total of individuals analysed. Differences in the TL distributions by season were assessed by applying the two-sample Kolmogorov-Smirnov test. Sex-ratio was calculated as the percentage of females regarding the total number of individuals in every 5 cm TL interval. These size intervals allowed us to have enough individuals to test for significant deviations from a 1:1 proportion of females and males using a chi-squared test. This analysis was used on the whole population sex-ratio and also by season. When significant differences were found, standardized residuals of the chi-squared test were analysed in order to identify which sex predominated using the criterion stablished by MacDonald and Gardner (2000)MacDonald P.L., Gardner R.C. 2000. Type I Error Rate Comparisons of Post Hoc Procedures for I j Chi-Square Tables. Educ. Psychol. Meas. 60: 735-754. https://doi.org/10.1177/00131640021970871 . This criterion sets an absolute value of ±1.96, from which the standardized residual is considered significant.
Allometric growth
⌅The length-weight relationships (TL-TW, TL-EW and DW-EW) and morphometric relationship (TL-DW) were established by fitting a potential curve to the data from the biological samplings of the commercial samples. For all these relationships, we used the same individuals in order to avoid variability. Differences in these relationships between males and females were tested using analysis of covariance (ANCOVA). To do so, lengths and weights were log-transformed in order to linearize their relationship. For length-weight relationships, weight was set as the dependent variable, while length and sex were set as continuous and categorical predictors, respectively. For the morphometric relationship, DW was set as the dependent variable, while TL and sex were set as continuous and categorical predictors, respectively. In case of non-significant differences between sexes, these relationships were recalculated for the whole population.
Reproduction
⌅The reproductive period was assessed from the seasonal evolution of the gonadosomatic index (GSI), calculated following
This index was only calculated with females in the stages “developing”, “spawning” and “capsule present” and with males in the stages “developing and “spawning”. Length at first maturity (at which 50% of the individuals are mature, L50) was calculated for females and males by fitting the logistic curve
where ML is the proportion of mature individuals for a given size class L and a and b are the parameters to be estimated. A log-likelihood function was used in SOLVER in EXCEL (Tokai 1997Tokai T. 1997. Maximum likelihood parameter estimates of a mesh selectivity logistic model through SOLVER on MS-Excel 1997. Bull. Jpn. Soc. Fish. Oceanogr. 61: 288-298.) to fit this model. Then, L50 was calculated as
Potential fecundity (PF), defined as the maximum number of yolked follicles observed in spawning-capable females (Maia et al. 2012Maia C., Erzini K., Serra-Pereira B., Figueiredo I. 2012. Reproductive biology of cuckoo ray Leucoraja naevus. J. Fish. Biol. 81: 1285-1296. https://doi.org/10.1111/j.1095-8649.2012.03412.x ), was considered in this work as the maximum number of vitellogenic oocytes (plus capsules if they were present) observed in a female. Also, the range of PF in all the analysed females was reported. Both PF and its range were determined for females in the “spawning” or “capsule present” maturity stages, both the maturity stages in which vitellogenic oocytes are present. The relationship between TL and PF was analysed by mean of linear regression of their log-transformed values. Mean PF values per season were also calculated.
Condition indices
⌅Different condition indices were calculated to evaluate their variation throughout the year in relation to reproduction. To this end, we used the data collected from the monthly samplings of the commercial bottom trawl fleet. The Le Cren index (k), which measures the deviation of an individual from the average weight for a given length (Lloret et al. 2014Lloret J., Shulman G., Love R.M. 2014. Condition and Health Indicators of Exploited Marine Fishes. John Wiley and Sons, New York, 262 pp. https://doi.org/10.1002/9781118752777 ), was calculated as
where W is the observed individual’s weight (g) and Wt , is the estimated weight of that fish obtained from the TL-W relationship. In the case of differences in the TL-W relationships between males and females, k was calculated for the males and females separately. The Le Cren index was calculated for both the TW and the EW. We also calculated the hepatosomatic (HSI) and digestivosomatic (DSI) indices as the percentage relative weight of the liver (LW) and digestive tract (SW) to EW, respectively. All these indices were calculated for all the individuals available.
Diet analysis
⌅All preys were identified to the lowest possible taxonomic level and counted, and their volume was measured using a trophometer, a calibrated device used in previous studies for this purpose (Olaso et al. 1998Olaso I., Velasco F., Perez N. 1998. Importance of discarded blue whiting (Micromesistius poutassou) in the diet of lesser spotted dogfish (Scyliorhinus canicula) in the Cantabrian Sea. ICES J. Mar. Sci. 55: 331-341. https://doi.org/10.1006/jmsc.1997.0263 , Valls et al. 2011Valls M., Quetglas A., Ordines F., Moranta J. 2011. Feeding ecology of demersal elasmobranchs from the shelf and slope off the Balearic Sea (western Mediterranean). Sci. Mar. 75: 633-639. https://doi.org/10.3989/scimar.2011.75n4633 ). The diet was described using the following indices: 1) frequency of occurrence (%F), calculated as the percentage of stomachs with a specific type of prey in relation to the total number of stomachs containing food; 2) numerical (%N) and volumetric (%V) composition, calculated as the percentage contribution of each prey to the whole content in number or volume, respectively; 3) index of relative importance, calculated using the formula, which was standardized as
according to Cortés (1997)Cortés E. 1997. A critical review of methods of studying fish feeding based on analysis of stomach. Can. J. Fish. Aquat. Sci. 54: 726-738. https://doi.org/10.1139/f96-316 ; and 4) the vacuity index, calculated as the percentage of empty stomachs.
Geographic and bathymetric distribution
⌅To assess the geographic and bathymetric distributions of R. polystigma, we used data collected during 832 hauls of the MEDITS surveys between 2003 and 2018. Individuals were classified into adults (above the L50) and juveniles (below the L50), and this sorting was used for both geographic and bathymetric distributions.
Maps of predicted abundance and persistence of adults and juveniles were created using the standardized abundance (n/km2) at each MEDITS sampling station. To this end, generalized additive models (GAMs) were used to predict the mean values in each of the 0.01×0.01° cell grids of the whole study area by year for juveniles and adults. The standardized abundances of the two population fractions were modelled using a two-stage GAM approach (Barry and Welsh 2002Barry S. C., Welsh A. H. 2002. Generalized additive modelling and zero inflated count data. Ecol. Modell. 157:179-188. https://doi.org/10.1016/S0304-3800(02)00194-1 ), owing to the presence of a larger proportion of zeros than expected from a Gaussian distribution (zero inflation). First, presence-absence was modelled with a logit link function and, second, the log-transformed abundance conditional on presence was modelled using an identity link function. The explanatory variables were year, included as a factor, and the smoothed effects of depth and the geographic position of the haul (latitude and longitude). Final model election was based on significance of the explanatory variables and minimization of the value of the Akaike information criterion. Finally, assuming independence of the two model stages, abundance predictions were obtained by multiplying the prediction of both stages. Although no specific correlation term was included in the modelling, the introduction of latitude and longitude as predictor variables implicitly incorporated the spatial structure. The mgcv package (Wood 2011Wood S. N. 2011. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stant. Soc. Ser. B Methodol. 73:3-36. https://doi.org/10.1111/j.1467-9868.2010.00749.x ) of the R statistical software (R Core Team 2018R Core Team. 2018. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available online at https://www.R-project.org/ ) was used for the whole modelling process.
Once the abundance maps were created, we identified the hotspot areas of each population fraction on a yearly basis following a similar approach to that of Colloca et al. (2015)Colloca F., Garofalo G., Bitetto I., et al. 2015. The Seascape of Demersal Fish Nursery Areas in the North Mediterranean Sea, a First Step Towards the Implementation of Spatial Planning for Trawl Fisheries. PLoS ONE 10: e0119590. https://doi.org/10.1371/journal.pone.0119590 . Initially, hotspot areas were identified as every grid cell of every particular year in which the predicted value was above the upper 95% confidence interval of the mean abundance for the whole time series. Second, hotspot persistence was defined as the percentage of years for which a particular cell was considered a hotspot. Third, we defined areas considered as persistent hotspots as those areas with persistence above 70%. Finally, spatial overlap of persistent hotspots of adults and juveniles was defined as cell grids in which both population fractions show persistent hotspots. Maps with the mean annual abundance and persistence of both population fractions and of the overlap of persistent hotspots were created using the Surfer 8 software.
To assess the bathymetric distribution, Generalized linear models (GLMs) and linear regression were used to evaluate the effect of depth (set as the continuous predictor) on the presence of the species at each sampling station and to obtain a bathymetric distribution model for each species. These models were applied for the whole population, the adult population and the juvenile population. The response type (unimodal or linear) of each fraction of the population was chosen first on the basis of resulting significant and second on the basis of the lowest value of the Akaike information criterion.
Population trend
⌅Frequency of appearance and annual abundance were used as indicators based on the surveys’ data to assess the population state of R. polystigma. We calculated the frequency of appearance and the annual abundance for each year (2003-2018) during the time series of the MEDITS surveys by dividing the number of sampling stations in which the species appeared by the total number of hauls carried out each year within the species’ bathymetric range and by calculating the mean annual individuals per km2, respectively. Baselines were calculated as the mean frequency of appearance and the mean annual abundance of the first three years (2003-2005).
We also assessed the evolution of the mean PF per km2 (population PF). To do so, we first applied the PF equation (which relates the TL and the number of oocytes produced) to estimate the fecundity (in number of produced oocytes) of each adult female (above the L50) in the MEDITS data. Then, using the swept area in the haul, we calculated the estimated number of oocytes produced per km2 in each haul. The average number of oocytes per km2 was calculated from all hauls carried out within the bathymetric range of distribution of R. polystigma, including those in which the species was not present (i.e. counting as zero). A linear regression model was used to detect significant temporal trends in frequency of appearance, abundance and population PF.
RESULTS
⌅Population structure
⌅Of the 236 individuals of R. polystigma, 98 were males and 138 females. The sizes ranged between 17 and 59 cm TL and between 11 and 39 cm DW. The species showed a size distribution with a mode between 37 and 39 cm (Fig. 1).
The size distribution changed throughout the year (Fig. 2). In the winter and summer seasons TL showed the widest range (between 16 and 60 cm TL in both seasons). In spring, the smallest and largest sizes did not appear and there was a mode in the sizes between 40 and 42 cm of TL. In summer, the mode was observed in sizes between 37 and 39 cm TL. In autumn, the mode was detected in 28 to 30 cm TL. Significant differences between the TL distributions by season resulting from the Kolmogorov-Smirnov test were found between summer and winter (p<0.05) and between autumn and spring (p<0.01).
The chi-squared test indicated that there were significant differences from the 1:1 sex proportion for the whole population (p=0.03), with a higher proportion of females. By size class, there were no significant differences except for the larger classes, in which females predominated among individuals larger than 40 cm (Fig. 3). Standardized residuals of the size classes 45 to 49 cm and 50 to 54 cm showed significant differences (z45-49= ±2.16, z50-54= ±2.07), indicating a higher proportion of females, whereas in the 55 to 60 cm class all the individuals were females. Analysed by season, no significant differences from the 1:1 sex proportion were found.
Allometric growth
⌅The ANCOVA analyses showed significant differences (p<0.05) between the two sexes for the relationships TL-TW, TL-EW and TL-DW. On the other hand, the DW-EW relationship showed no significant differences between males and females (p=0.17). Except for the TL-DW relationship, the remaining allometric coefficients (b) of the tested regressions were greater than 3 (Table 2), indicating a positive allometry growth.
| Relationship tested | Sex | a | b | R2 |
|---|---|---|---|---|
| TL-TW | F | 0.0006 | 3.62 (3.52-3.72) | 0.97 |
| M | 0.0012 | 3.38 (3.26-3.49) | 0.98 | |
| TL-EW | F | 0.0005 | 3.62 (3.51-3.71) | 0.97 |
| M | 0.001 | 3.4 (3.28-3.52) | 0.98 | |
| DW-EW | F | 2·10-6 | 3.37 (3.23-3.45) | 0.97 |
| M | 3·10-6 | 3.34 (3.26-3.41) | 0.99 | |
| whole population | 3·10-6 | 3.33 (3.24-3.38) | 0.97 | |
| TL-DW | F | 0.4694 | 1.05 (1.03-1.08) | 0.98 |
| M | 0.5921 | 1.01 (0.98-1.04) | 0.97 |
Reproduction
⌅The size ranges (TL) observed for every maturity stage in females were 20 to 41 cm for immature individuals, 27 to 45 cm for developing individuals, 40 to 51 cm for spawning individuals, 44 to 58 cm for capsule present individuals and 40 to 59 cm for regenerating individuals. In males, the size ranges were 17 to 39 cm for immature individuals, 33 to 44 cm for developing individuals and 37 to 52 cm for spawning individuals.
Females showed a maximum GSI in summer (Fig. 4), whereas in winter they showed the maximum proportion of spawning individuals and the maximum proportion of individuals with egg capsules present. Egg capsules were also present in spring, but they did not appear in summer and autumn. In summer, all females showed a developing or regenerating stage. The proportion of regenerating stage was highest in the summer and autumn months. The GSI values for males were the highest and very similar in autumn, winter and spring, whereas a the lowest GSI was observed in summer. The highest proportion of spawning males was found in winter. In the rest of the seasons, proportions of spawning and developing individuals were similar (Fig. 4). The proportion of immatures by season was 23% in both winter and spring, 52% in summer and 28% in autumn for females and 47% in winter, 26% in spring, 54% in summer and 40% in autumn for males.
Females showed between 11 and 45 yolked oocytes per individual (PF = 45), with a significant and positive relationship between PF and TL (p=0.008) (Fig. 5). Mean values of PF±SE per season were 18.83±26.64, 12.72±55.06 and 23.57±67.5, for autumn, spring and winter, respectively.
The smallest sexually active individuals found measured 41 and 33 cm for females and males, respectively. The L50 was also larger for females (47.7 cm) than for males (38.4 cm) (Fig. 6).
Condition indices
⌅Immature individuals showed an annual cycle based on condition indices (Fig. 7). Their HSI and k followed the same pattern, showing minimum values in winter and continuous recovery until their maximum values in summer. Their DSI pattern showed its maximum in spring and then it started to decrease to a minimum in winter.
Mature males and females showed a similar seasonal pattern of variation of condition indices. For both sexes, the Le Cren index (k) showed maximum values in autumn and winter for both sexes, an abrupt decrease in spring and a recovery from then on (Fig. 7). The pattern for HSI and DSI was different from that for k, but again similar between males and females (Fig. 7), with a maximum during the summer months. After summer, values started decreasing until winter (in males) and spring (in females), and then they started to increase again until a peak in summer.
Diet
⌅A total of 152 stomachs of individuals ranging from 18 to 58 cm TL were analysed. Of these, 12 were empty (vacuity index =7.9%). The main taxonomic groups of prey were decapod crustaceans, which were present in 57% of the stomachs representing 56% of the index of relative importance (IRI), fish, which were present in 19% of stomachs representing 19%IRI and amphipods, which were present in 31% of stomachs representing 12%IRI. Isopods, lophogastrids, mysidaceans and annelids were also frequently found in the stomach contents (17-26% of the stomachs), but with lower %IRI (around 3%IRI). Molluscs and cnidarians completed the diet with low frequencies and %IRI (Table 3).
| Prey | F | %N | %V | IRI | %IRI | |
|---|---|---|---|---|---|---|
| Decapoda | 57.14 | 36.67 | 26.66 | 3618.85 | 55.91 | |
| Natantia | 27.86 | 9.52 | 3.58 | 364.97 | 11.71 | |
| Processa spp. | 15.00 | 6.92 | 3.34 | 153.91 | 4.94 | |
| Solenocera membranacea | 12.14 | 3.34 | 8.56 | 144.45 | 4.63 | |
| Liocarcinus spp. | 7.86 | 2.22 | 4.69 | 54.30 | 1.74 | |
| Philocheras sculptus | 2.86 | 6.30 | 0.52 | 19.49 | 0.63 | |
| Brachyura | 5.00 | 1.11 | 0.99 | 10.53 | 0.34 | |
| Portunidae | 2.14 | 0.99 | 0.91 | 4.06 | 0.13 | |
| Crangonidae | 1.43 | 2.10 | 0.64 | 3.92 | 0.13 | |
| Philocheras spp. | 1.43 | 2.10 | 0.14 | 3.21 | 0.10 | |
| Processa canaliculata | 2.86 | 0.49 | 0.42 | 2.62 | 0.08 | |
| Plesionika antigai | 0.71 | 0.12 | 1.94 | 1.48 | 0.05 | |
| Galathea intermedia | 1.43 | 0.99 | 0.03 | 1.46 | 0.05 | |
| Alpheus spp. | 0.71 | 0.12 | 0.40 | 0.38 | 0.01 | |
| Pagurus spp. | 0.71 | 0.12 | 0.35 | 0.34 | 0.01 | |
| Scyllarus pygmaeus | 0.71 | 0.12 | 0.13 | 0.18 | 0.01 | |
| Pandalidae | 0.71 | 0.12 | 0.01 | 0.10 | <0.01 | |
| Fish | 19.29 | 8.40 | 55.57 | 1233.59 | 19.06 | |
| Gymnammodytes cicerelus | 17.86 | 6.80 | 39.17 | 820.84 | 26.33 | |
| Pisces | 7.14 | 1.11 | 5.49 | 47.14 | 1.51 | |
| Glossanodon leioglossus | 0.71 | 0.12 | 4.42 | 3.25 | 0.10 | |
| Gadiculus argenteus | 0.71 | 0.12 | 3.64 | 2.69 | 0.09 | |
| Trachurus trachurus | 0.71 | 0.12 | 2.68 | 2.00 | 0.06 | |
| Pseudaphya ferreri | 0.71 | 0.12 | 0.24 | 0.26 | 0.01 | |
| Amphipoda | 31.43 | 23.46 | 1.36 | 780.03 | 12.05 | |
| Gammaridae | 31.43 | 22.25 | 1.10 | 733.76 | 23.54 | |
| Amphipoda | 1.43 | 1.11 | 0.16 | 1.82 | 0.06 | |
| Hyperiidae | 0.71 | 0.12 | 0.07 | 0.14 | 0.00 | |
| Isopoda | 21.43 | 7.16 | 4.50 | 249.94 | 3.86 | |
| Isopoda | 21.43 | 6.06 | 3.86 | 212.41 | 6.81 | |
| Cirolanidae | 0.71 | 0.99 | 0.60 | 1.14 | 0.04 | |
| Eurydice spp. | 0.71 | 0.12 | 0.04 | 0.12 | <0.01 | |
| Lophogastrida | 26.43 | 6.30 | 2.12 | 222.33 | 3.43 | |
| Lophogaster typicus | 26.43 | 6.30 | 2.11 | 222.28 | 7.13 | |
| Mysida | 17.14 | 10.00 | 0.55 | 180.85 | 2.79 | |
| Annelida | 21.43 | 4.20 | 4.22 | 180.48 | 2.79 | |
| Polychaeta | 15.71 | 2.97 | 2.01 | 78.14 | 2.51 | |
| Annelida | 5.71 | 1.24 | 2.20 | 19.66 | 0.63 | |
| Other crustaceans | 5.00 | 2.96 | 0.32 | 16.41 | 0.25 | |
| Unidentified Crustacea | 4.29 | 2.84 | 0.31 | 13.50 | 0.43 | |
| Cumacea | 0.71 | 0.12 | 0.01 | 0.10 | <0.01 | |
| Mollusca | 2.86 | 0.49 | 2.47 | 8.48 | 0.13 | |
| Loligo forbesii | 0.71 | 0.12 | 2.23 | 1.68 | 0.05 | |
| Cephalopoda | 1.43 | 0.25 | 0.24 | 0.69 | 0.02 | |
| Cavolinia inflexa | 0.71 | 0.12 | 0.01 | 0.09 | <0.01 | |
| Cnidaria | 0.71 | 0.12 | 1.89 | 1.44 | 0.02 | |
| Pelagia noctiluca | 0.71 | 0.12 | 1.90 | 1.44 | 0.05 | |
| Unidentified remains | 0.71 | 0.12 | 0.28 | 0.29 | <0.01 | |
| Algae | 0.71 | 0.12 | 0.05 | 0.12 | <0.01 | |
Within decapods, 16 prey items were identified, unidentified natantians being the main prey, representing 11.7 of %IRI. Processa spp and Solenocera membranacea were among the identified species showing the highest frequencies (15% and 12%, respectively) and %IRI (4.9 and 4.6, respectively). Among fish, Gymnammodytes cicerelus, despite not being as frequent as the previous species, was the most important prey in terms of %IRI (26%) owing to its larger percentage of volume (39%).
Geographic and bathymetric distribution
⌅The mean predicted values of abundance and the persistence of hotspots showed that individuals are mainly found off the northwest coast of Mallorca, in the Menorca Channel and off the east coast of Menorca (Fig. 8). This distribution pattern was similar for both adults and juveniles, although the latter showed higher abundances and a deeper distribution. However, on the northwest coast of Mallorca, high abundances and persistence of hotspots of juveniles were also detected on the shallow continental shelf. This area and the east of Menorca were the only ones where persistent hotspots for adults and juveniles overlapped (Fig. 8).
Raja polystigma individuals were captured between 46 and 375 m depth. GLMs showed a better fit to a unimodal bathymetric distribution model for the whole (p<0.05) and juvenile (p<0.001) populations, whereas the adult population was better adjusted to a linear regression (p<0.001).
Adults and juveniles showed bathymetric segregation. Adults were related to the continental shelf (<200 m depth), with their maximum of occurrence at 50 m depth, whereas juveniles were found from the shelf to the upper slope, and showed their optimum depth on the shelf break around 160 m depth (Fig. 9).
Population trend
⌅Frequency of appearance showed no significant trend, as it remained stable throughout the time series analysed (Fig. 10, p=0.135). However, all the values remained above the baseline in the whole time series. Annual abundance showed no significant trend either. All of its values remained above the baseline, except for 2009 (Fig. 10, p=0.18). Furthermore, the population PF showed no significant trend either (Fig. 11, p=0.058).
DISCUSSION
⌅Despite the critical situation of almost all Mediterranean batoid populations, little is known about their basic biological parameters. The present study estimated essential biological parameters (population structure, growth, diet, somatic condition, reproduction and geographic and bathymetric distribution) of Raja polystigma, an endemic species in the Mediterranean.
The analysed population of R. polystigma showed similar maximum lengths (around 60 cm TL) to those reported for this species in the central Mediterranean (Porcu et al. 2020Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 ), around 60 cm TL. Raja polystigma probably reproduces throughout the year in the Balearic Islands, as males in the “spawning” stage and “regenerating” females have been observed throughout the year. However, a peak was observed in the spawning activity in winter and spring. The highest percentage of individuals with egg capsules was detected in winter, and maximum GSI values of females were observed in spring, suggesting that reproduction in this species occurs mainly during the first half of the year in the study area. This finding is similar to that of Capapé and Quignard (1978)Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244., who also report the winter as the main reproduction period. However, Porcu et al. (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 observed that the reproduction period in individuals collected off Sardinia was mainly in summer. The spawning peak in winter and spring in the Balearic Islands is also coincident with the seasonal variations in the length-frequency distributions. The highest number of egg capsules was recorded in winter, and was followed in spring by a cohort of recruits, individuals of 16 to 25 cm TL, which corresponds to the TL reported for new-borns by Porcu et al (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 . These recruits were not detected in summer but could correspond to the mode of 26 to 28 cm TL detected in autumn.
The sex ratio showed a significantly higher proportion of females in individuals larger than 45 cm TL. These differences could be due to the greater longevity reported in females as opposed to a faster growth, because males and females show similar growth rates, k = 0.32 and k = 0.28, respectively (Porcu et al. 2020Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 ).
Males and females showed significant differences in their length-weight relationships, females being heavier than males at a given TL. These differences cannot be related to the gonad weight since the same results were also found when the length-eviscerated weight relationship was analysed. According to Orlov and Cotton (2011)Orlov A.M., Cotton C.F. 2011. Sexually Dimorphic Morphological Characters in Five North Atlantic Deepwater Skates (Chondrichthyes: Rajiformes). J. Mar. Biol. 1-18. https://doi.org/10.1155/2011/842821 differences in body shape caused by sexual dimorphism could be the reason for the observed differences in weight by sex.
The L50 values of R. polystigma showed a difference of 9 cm TL between females (47.7 cm TL) and males (38.4 cm TL). These differences are lower than those reported in the central Mediterranean, where L50 is estimated at 50.6 and 48.8 cm TL for females and males, respectively (Porcu et al. 2020Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 ). A similar pattern in size differences at maturity related to sex has been observed in other skate species such as Raja brachyura from Sardinia waters (L50=87 cm for females and 81 cm for males; Porcu et al. 2015Porcu C., Bellodi A., Cannas R., et al. 2015. Life-history traits of the commercial blonde ray, Raja brachyura, from the central-western Mediterranean Sea. Mediterr. Mar. Sci. 16: 90-102. https://doi.org/10.12681/mms.898 ) and Raja clavata from Tunisia (L50=81 cm for females and 67 cm for males; Kadri et al. 2014Kadri H., Marouani S., Saïdi B., et al. 2014. Age, growth, sexual maturity and reproduction of the thornback ray, Raja clavata (L.), of the Gulf of Gabès (south-central Mediterranean Sea). Mar. Biol. Res. 10: 416-425. https://doi.org/10.1080/17451000.2013.797584 ) and from the Balearic Islands (L50=76.4 cm for females and 67.3 cm for males; Ramírez-Amaro et al. 2020Ramírez-Amaro S., Ordines F., Esteban A., et al. 2020. The diversity of recent trends for chondrichthyans in the Mediterranean reflects fishing exploitation and a potential evolutionary pressure towards early maturation. Sci. Rep. 10: 547. https://doi.org/10.1038/s41598-019-56818-9 ).
PF of R. polystigma increases with size, as previously suggested by Capapé and Quignard (1978)Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244.. The PF estimated here ranged between 11 and 45 yolked follicles per individual per year. A wide range of between 20 and 62 yolked follicles per year was also reported by Capapé and Quignard (1978)Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244.. In contrast, Porcu et al. (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 reported a shorter range of between 14 and 19 yolked follicles per year. Means of PF by season matched with the estimated reproductive season (winter-spring), because the maximum values of PF were found in winter and autumn, just before the detected egg-laying episodes. However, the presence of females in the “regenerating” stage throughout the year and the unknown duration of this stage indicate that egg-laying females could be present throughout the year. They could therefore also be present in summer, a season during which we did not detect them, although regenerating females were found during autumn.
In view of the results obtained in the condition indices, we can consider R. polystigma as a capital breeder. Capital breeders are species that acquire resources in advance and store them until they need them for reproductive investment (Jönsson 1997Jönsson K. I. 1997. Capital and Income Breeding as Alternative Tactics of Resource Use in Reproduction. Oikos 78: 57-66. https://doi.org/10.2307/3545800 ). Hepatosomatic index and DSI were used in this study as approximate measures of the energy stores because of the important role played by the liver in lipid storage (Lloret et al. 2014Lloret J., Shulman G., Love R.M. 2014. Condition and Health Indicators of Exploited Marine Fishes. John Wiley and Sons, New York, 262 pp. https://doi.org/10.1002/9781118752777 ) and food intake, respectively. Mature individuals showed an increase in HSI and DSI during summer and a decrease afterwards during their spawning season (winter). This may indicate that individuals of R. polystigma increase their feeding activity when preparing for spawning and, as a result, the resources stored in the liver increase so that they can be finally used during the spawning season. This also happens in other elasmobranchs, such as Mobula eregoodoo, although increased food intake has also been observed during reproduction (Lawson et al. 2022Lawson C.L., Dudgeon C.L., Richardson A.J., et al. 2022. Flexibility for fuelling reproduction in a pelagic ray (Mobula eregoodoo) suggested by bioenergetic modelling. J. Fish Biol. 1-10. https://doi.org/10.1111/jfb.14995 ).
Regarding the diet of this skate, our results suggest that the main prey of R. polystigma are decapod crustaceans, specially natantians, in agreement with previous reported studies in the Balearic Islands (Valls et al. 2011Valls M., Quetglas A., Ordines F., Moranta J. 2011. Feeding ecology of demersal elasmobranchs from the shelf and slope off the Balearic Sea (western Mediterranean). Sci. Mar. 75: 633-639. https://doi.org/10.3989/scimar.2011.75n4633 ). Nonetheless, it is also a highly piscivorous species. Among fish prey, R. polystigma showed a preference for a single species (G. cicerelus, representing 26%IRI), whose behaviour consists of burrowing itself completely in the sand sediment (Wheeler 1979Wheeler A. 1979. Ammodytidae. In: Hureau J. C., Monod T. (eds), Check-list of the fishes of the north-eastern Atlantic and of the Mediterranean (CLOFNAM). UNESCO, pp. 446-448.). This could make it more accessible to R. polystigma than other fishes that are less in contact with the substrate. Also, the fact that G. cicerelus, S. membranacea (Demestre and Abelló 1993Demestre M., Abelló P. 1993. Growth and distribution of Solenocera membranacea (Risso, 1816) (Decapoda, Dendrobranchiata) in the northwestern Mediterranean Sea. Sci. Mar. 57: 161-166., Cartes et al. 2007Cartes J.E., Huguet C., Parra S., Sanchez F. 2007. Trophic relationships in deep-water decapods of Le Danois bank (Cantabrian Sea, NE Atlantic): Trends related with depth and seasonal changes in food quality and availability. Deep-Sea Res. I: Oceanogr. Res. Pap. 54: 1091-1110. https://doi.org/10.1016/j.dsr.2007.04.012 ) and Lophogaster typicus (Cartes et al. 2011Cartes J.E., Mamouridis V., Fanelli E. 2011. Deep-sea suprabenthos assemblages (Crustacea) off the Balearic Islands (western Mediterranean): Mesoscale variability in diversity and production. J. Sea Res. 65: 340-354. https://doi.org/10.1016/j.seares.2011.02.002 ) are among its most frequent prey might indicate a preference for muddy and sandy-muddy bottoms for feeding.
Adults of R. polystigma are mainly found on the continental shelf from 45 to 180 m depth, whereas juveniles are mainly found on the deep shelf from 50 to 360 m depth, with an optimum depth of around 160 m, which agrees with the finding by Porcu et al. (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 . It also agrees with the general bathymetric distribution described in Tunisia (from 50 to 400 m depth; Capapé and Quignard, 1978Capapé C., Quignard J.P. 1978. Contribution a la biologie des Rajidae des côtes tunisiennes. XIV - Raja polystigma Regan, 1923. Répartition géographique et bathymétrique, sexualité, reproduction, fécondité. Cah. Biol. Mar. 19: 233-244.) and in Sicily (Geraci et al. 2021Geraci M.L., Ragonese S., Scannella D., et al. 2021. Batoid abundances, spatial distribution, and life history traits in the strait of Sicily (Central Mediterranean Sea): Bridging a knowledge gap through three decades of survey. Animals 11: 2189. https://doi.org/10.3390/ani11082189 ). However, Porcu et al. (2020)Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 suggest that R. polystigma completes its life cycle in shallow waters. Although our data are limited to samplings carried out from 50 m depth to the slope, the abundance of R. polystigma decreases at depths of less than 50 m in the Balearic Islands, where bottoms are dominated by Posidonia oceanica. Raja polystigma is known to mainly inhabit soft bottoms (Porcu et al. 2020Porcu C., Bellodi A., Cau A., et al. 2020. Uncommon biological patterns of a little known endemic Mediterranean skate, Raja polystigma (Risso, 1810). Reg. Stud. Mar. Sci. 34: 101065. https://doi.org/10.1016/j.rsma.2020.101065 ). Moreover, catches of R. polystigma from the artisanal fleet operating shallower than 50 m depth in the Balearic Islands are negligible (pers. comm.).
The bathymetric segregation of adults and juveniles may avoid intraspecific competition. This kind of segregation has also been observed in other elasmobranch species, such as Scyliorhinus canicula (Massutí and Moranta 2003Massutí E., Moranta J. 2003. Demersal assemblages and depth distribution of elasmobranchs from the continental shelf and slope off the Balearic Islands (western Mediterranean). ICES J. Mar. Sci. 60: 753-766. https://doi.org/10.1016/S1054-3139(03)00089-4 ). Nonetheless, this segregation often happens the other way around. For instance, juveniles of the shark Galeus melastomus are found shallower, on the upper slope (300-500 m), than adults, which inhabit the middle slope (500-700 m) (Massutí and Moranta 2003Massutí E., Moranta J. 2003. Demersal assemblages and depth distribution of elasmobranchs from the continental shelf and slope off the Balearic Islands (western Mediterranean). ICES J. Mar. Sci. 60: 753-766. https://doi.org/10.1016/S1054-3139(03)00089-4 ). The deep shelf and the upper slope have recently been described as a hotspot of recruitment and community nekto-benthic species richness and density in the Mediterranean (Tugores et al. 2019Tugores M.P., Ordines F., Guijarro B., et al. 2019. Essential fish habitats and hotspots of nekto-benthic diversity and density in the western Mediterranean. Aquat. Conserv.: Mar. Freshw. Ecosyst. 29: 461-471. https://doi.org/10.1002/aqc.3031 ). This seems also to be the case of R. polystigma, for which the deep shelf, shelf break and beginning of the upper slope are very important areas where recruits aggregate and may represent an essential fish habitat for this species.
Raja polystigma seems to be widely distributed geographically in the study area and also occupies a wide bathymetric range. Our results show that it is widespread around the islands of Mallorca and Menorca. The whole study area seems suitable for this species and what really determines its abundance is the availability of its preferred bathymetric range (from 45 m to 400 m depth). However, fishing activity might also exert some effect on their abundance. Elasmobranchs are particularly vulnerable to fishing pressure (Stevens et al. 2000Stevens J.D., Bonfil R., Dulvy N.K., Walker P.A. 2000. The effects of fishing on sharks, rays, and chimaeras (chondrichthyans), and the implications for marine ecosystems. ICES J. Mar. Sci. 57: 476-494. https://doi.org/10.1006/jmsc.2000.0724 ). The Balearic Islands show both lower levels of fishing exploitation and higher diversity of demersal elasmobranchs than adjacent waters off the Iberian Peninsula (Massutí and Moranta 2003Massutí E., Moranta J. 2003. Demersal assemblages and depth distribution of elasmobranchs from the continental shelf and slope off the Balearic Islands (western Mediterranean). ICES J. Mar. Sci. 60: 753-766. https://doi.org/10.1016/S1054-3139(03)00089-4 ). Our data seem to indicate that fishing exploitation also affects the abundance distribution of R. polystigma in the Balearic Islands, since although the southeast and southwest of Mallorca have its preferred bathymetric range, this area is subjected to higher exploitation than other areas in the archipelago (Farriols et al. 2017Farriols M.T., Ordines F., Somerfield P.J., et al. 2017. Bottom trawl impacts on Mediterranean demersal fish diversity: Not so obvious or are we too late? Cont. Shelf Res. 137: 84-102. https://doi.org/10.1016/j.csr.2016.11.011 ), and consequently the abundances are lower.
Raja polystigma has never been assessed previously in the Balearic Islands. Because of its low abundance and the fact it is not a target species, information on the size structure of the catches is scarce, so structural models cannot be used to assess its state of exploitation. Therefore, we conducted a preliminary assessment of this species based on the trends of its frequency of appearance, its annual abundance and its PF in the Balearic Islands, which resulted in stable trends. Stability has also been reported for other skate species in the area, R. miraletus, R. radula, Leucoraja naevus and Dipturus oxyrinchys, and even recovery in the case of R. clavata (Ramírez-Amaro et al. 2020Ramírez-Amaro S., Ordines F., Esteban A., et al. 2020. The diversity of recent trends for chondrichthyans in the Mediterranean reflects fishing exploitation and a potential evolutionary pressure towards early maturation. Sci. Rep. 10: 547. https://doi.org/10.1038/s41598-019-56818-9 ). In addition, from commercial catch data, Guijarro et al. (2012)Guijarro B., Quetglas A., Moranta J., et al. 2012. Inter- and intra-annual trends and status indicators of nektobenthic elasmobranchs off the Balearic Islands (northwestern Mediterranean). Sci. Mar. 76: 87-96. https://doi.org/10.3989/scimar.03432.22A observed that elasmobranchs showed a stable trend or even some recovery on the shelf of the Balearic Islands. The stable trend could be related to the reduction of the bottom trawl fishing effort on the shelf and its transfer to greater depths (Quetglas et al. 2016Quetglas A., Merino G., González J., et al. 2016. Regional Implementation Plan for Demersal Fisheries from the Balearic Islands (Western Mediterranean). Technical report in the framework of MYFISH project (Maximising yield of fisheries while balancing ecosystem, economic and social concerns. Grant Agreement No: 289257)., Farriols et al. 2019Farriols M.T., Ordines F., Carbonara P., et al. 2019. Spatio-temporal trends in diversity of demersal fish assemblages in the Mediterranean. Sci. Mar. 83S1: 189-206. https://doi.org/10.3989/scimar.04977.13A , Ramírez-Amaro et al. 2020Ramírez-Amaro S., Ordines F., Esteban A., et al. 2020. The diversity of recent trends for chondrichthyans in the Mediterranean reflects fishing exploitation and a potential evolutionary pressure towards early maturation. Sci. Rep. 10: 547. https://doi.org/10.1038/s41598-019-56818-9 ).